Predictive Control Method For Robot Teleoperation With Time Stamp

Teleoperation robot has been deep into all aspects of society. For example, in modern industrial production teleoperation technology has been widely used in remote manufacturing, telemedicine, and other hazardous environments job. Teleoperation robot not only can reduce some dangerous activities, but also greatly improve the efficiency of work. But the phenomenon of the network which has random delay, packet loss and reordering phenomenon has seriously affected the performance of teleoperation systems, and even cause system instability. Therefore, how to solve these problems becomes the emphasis and difficulty in the field of network teleoperation robot.Aiming at the problem of network robot system time delay control, this paper presents a predictive control method for robot teleoperation with timestamp. First, we bulid the structure prediction with time stamp telerobot control system based on analyzing the data flow of the operating system. Second, we collect and analysis the one-way delay jitter on data based on writing a one-way network delay jitter test program. Finally, we build the system which uses the force feedback control as the host controller and uses the OpenGL environment to establish the virtual IRB120 robot and use the autoregressive model, sparse multivariate linear regression and particle swarm optimization (pso) to predict the experiment. The experimental results show that predictive control method for robot teleoperation with timestamp is a effective way to deal with the network time delay problem, the control effect which adopts the method of sparse multivariate linear regression is better.